As of January 16, 1998, we have recruited 138 stroke patients and TIA patients into the study. 616 patients have been assessed for inclusion, representing an inclusion frequency of 23%. The principle reason for refusal to participate in the study was a lack of interest on the part of the patient or relatives or a premature discharge from hospital before assessment could take place. This latter usually meant that patients were discharged by attending physicians other than those in the Cerebrovascular Program within 24 hours of admission. The enrolled patients are generally representative of those not enrolled in the study. More african americans were enrolled in the study than white individuals, representing the racial distribution in the Baltimore area. An equal number of male and female patients were recruited, which once again is representative of the nonenrolled group. As before, the principle reasons for non-enrollment in the study are significant obtundation, significant concomitant disease and radiological evidence of previous stroke. The majority of patients included in this study suffered a hemispheric infarct. Approximately 7% of patients had a hemispherical brain stem hemorrhage. This is consistent with the incidence of cerebral hemorrhage in the general population. The distribution and nature of strokes is similar in the patients included in the study and those not enrolled in the study, once again emphasizing that the study population is representative of the general population of stroke patients admitted to the hospital. Left and right hemispheric and brain stem strokes are equally represented without laterality predominance. The distribution of stroke is similar in those patients not enrolled in the study as compared to those enrolled in the study, once again emphasizing the lack of bias. Nearly 100% of patients underwent cardiac autonomic testing, Holter tape assessment and late potential analysis. Transesophageal echocardiography was performed in 76% of patients in the study, but not performed in 24%, either because this was refused by the patient or was considered to be contraindicated by the cardiology staff, or because these patients had had a recent transesophageal echocardiogram. Adenosine thallium scanning was performed in nearly 60% of the patients. In those patients in whom it was not performed, there was a documented history of coronary artery disease. Approximately 30% of patients did not attend a follow-up assessment.